What #6 on Hacker News means

Last week I posted an article on my experiences at the Wisdom 2.0 conference It reached #6 on Hacker News, far surpassing anything else I have written. I posted this article to HN, Twitter and Reddit, but I have only 1 karma point on Reddit, and I have not made a real post on Twitter in two years, so almost all of this traffic is purely from HN. Below I am offering some data and screenshots from Google Analytics to show what reaching this level on HN produced in traffic to my blog, which might be a good comparison against a similar post made by 37Signals recently. Perhaps my blogging-fu is so powerful that the post rose organically, but I did have a simple strategy for posting that I used to try to maximize the impact. And, I have a question on how ranking degradation occurs once a link is in HN that perhaps you can answer.

My strategy was simple: I posted the article at 11 am EST. I thought this window would optimize for getting people commuting into work on the West Coast, slide into the lunch time browsing of people on the East Coast, and get people at the end of their day in Europe. I wanted to avoid the dead times in the middle of the work day when paranoid project managers with low blood sugar are constantly peeking over the shoulders of their employees. Perhaps my strategy was confirmed, as most of the traffic was from the coastal US. Unexpectedly, there was lots of traffic from Texas, which could have been from attendees of the interactive SXSW conference who on that day only did not have their bosses watching over their shoulders telling them to get back to work.

Over 7,000 hits that day. Much to my dismay, an hour after posting the link into HN I saw Google Analytics was not installed (theme upgrade broke the GA plugin, doh!), so I imagine I missed at least several hundred hits in GA.

Peaking at 160 active visitors. Watching Google Analytics Real-Time is a fascinating way to watch the ebbs and flows of progress in once you send a post out into the wild. Almost all the traffic was from HN, or derivative sites.

My 1.5 GB Slicehost server never got above 20% CPU. Surprisingly, apache load was heavy and I almost never saw varnish in top. Isn’t varnish supposed to remove that load on apache? Perhaps I misconfigured it (varnishd -f /etc/varnish/default.vcl -s malloc,1G -T -a

My biggest question: the next day there were several stories on the top three pages of HN but with fewer points and that had been posted around the same time. My story had drifted out by noon the following day, which seemed strange. I suspect the aging algorithm on HN includes commenting frequency; I had only 17 comments in total, while other stories were generating more discussion.